Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method comprising: (a) wirelessly transmitting, by orthogonal frequency-division multiplexing (OFDM), a largeband signal in such a way that the largeband signal travels from one or more transmitters to a backscatter node, reflects from the backscatter node and travels to one or more RF receivers; (b) taking measurements, with the one or more RF receivers, of the largeband signal; (c) identifying, based on the measurements, (i) when the backscatter node is in a first reflective state and when the backscatter node is in a second reflective state, the first reflective state differing from the second reflective state in that a reflection response of the backscatter node is different in the first reflective state than in the second reflective state, which reflection response is phase or amplitude of RF reflection from the backscatter node as a function of RF radiation incident on the backscatter node, or (ii) when the backscatter node is transitioning between the first and second reflective states; (d) calculating estimates of phase or amplitude of the largeband signal for each frequency band in a set of multiple frequency bands of the largeband signal, which estimates are based on a portion of the measurements, which portion of the measurements is taken when the backscatter node is in the first reflective state or in the second reflective state and is not taken when the backscatter node is transitioning between the first and second reflective states; and (e) calculating, for each specific frequency band in the set of frequency bands, an average phase and average amplitude, by subtracting a first average phase from a second average phase and by subtracting a first average amplitude from a second average amplitude, where (i) the first average phase and first average amplitude are average phase and average amplitude, respectively, for the specific frequency band as measured by the one or more receivers during OFDM symbols that occur only while the backscatter node is in a non-reflective state, and (ii) the second average phase and second average amplitude are average phase and average amplitude, respectively, for the specific frequency band as measured by the one or more receivers during OFDM symbols that occur only while the backscatter node is in a reflective state.
This invention relates to wireless communication systems using orthogonal frequency-division multiplexing (OFDM) to detect and analyze backscatter signals from a node that can switch between reflective states. The problem addressed is accurately measuring phase and amplitude changes in reflected signals to determine the state of the backscatter node, which may be used for sensing or communication purposes. The method involves transmitting a wideband OFDM signal from one or more transmitters to a backscatter node, which reflects the signal to one or more RF receivers. The receivers measure the reflected signal to identify when the backscatter node is in a first reflective state, a second reflective state, or transitioning between them. The reflective states differ in their phase or amplitude response to incident RF radiation. The system calculates phase and amplitude estimates for each frequency band in the OFDM signal, using measurements taken only when the node is in a stable reflective state, excluding transition periods. For each frequency band, the method computes an average phase and amplitude difference by subtracting measurements taken during non-reflective states from those taken during reflective states. This allows precise determination of the backscatter node's state based on signal characteristics. The technique is useful for applications requiring high-resolution sensing or low-power communication using backscatter nodes.
2. A method comprising: (a) wirelessly transmitting a largeband signal in such a way that the largeband signal travels from one or more transmitters to a backscatter node, reflects from the backscatter node and travels to one or more RF receivers; (b) taking measurements, with the one or more RF receivers, of the largeband signal; (c) identifying, based on the measurements, (i) when the backscatter node is in a first reflective state and when the backscatter node is in a second reflective state, the first reflective state differing from the second reflective state in that a reflection response of the backscatter node is different in the first reflective state than in the second reflective state, which reflection response is phase or amplitude of RF reflection from the backscatter node as a function of RF radiation incident on the backscatter node, or (ii) when the backscatter node is transitioning between the first and second reflective states; and (d) calculating estimates of phase or amplitude of the largeband signal for each frequency band in a set of multiple frequency bands of the largeband signal, which estimates are based on a portion of the measurements, which portion of the measurements is taken when the backscatter node is in the first reflective state or in the second reflective state and is not taken when the backscatter node is transitioning between the first and second reflective states; wherein the transmitting of the largeband signal involves transmitting a set of orthogonal frequency-division multiplexing (OFDM) symbols in such a way that (i) each OFDM symbol has a duration, and (ii) the duration is sufficiently short that each OFDM symbol in a subset of the set of OFDM symbols (A) occurs only while the backscatter node remains in the first reflective state or remains in the second reflective state, and (B) does not occur while the backscatter node is transitioning between the first and second reflective states.
This invention relates to wireless communication systems using backscatter nodes for signal reflection. The problem addressed is accurately measuring phase or amplitude of a wideband signal reflected by a backscatter node that switches between different reflective states. The solution involves transmitting a wideband signal using orthogonal frequency-division multiplexing (OFDM) symbols with durations short enough to ensure each symbol is captured while the backscatter node remains in a stable reflective state, avoiding transitions. The system includes one or more transmitters sending the signal to the backscatter node, which reflects it to one or more RF receivers. The receivers measure the reflected signal to identify when the backscatter node is in a first or second reflective state, where the states differ in phase or amplitude response. The system calculates phase or amplitude estimates for each frequency band of the wideband signal using measurements taken only during stable reflective states, excluding transition periods. The OFDM symbols are structured to ensure each symbol aligns with a stable state, improving measurement accuracy. This approach enables precise signal characterization despite the backscatter node's state changes.
3. A method comprising: (a) wirelessly transmitting a largeband signal in such a way that the largeband signal travels from one or more transmitters to a radio frequency identification (RFID) tag, reflects from the RFID tag and travels to one or more RF receivers; (b) taking measurements, with the one or more RF receivers, of the largeband signal; (c) identifying, based on the measurements (i) when the RFID tag is in a first reflective state and when the RFID tag is in a second reflective state, the first reflective state differing from the second reflective state in that a reflection response of the RFID tag is different in the first reflective state than in the second reflective state, which reflection response is phase or amplitude of RF reflection from the RFID tag as a function of RF radiation incident on the RFID tag, or (ii) when the RFID tag is transitioning between the first and second reflective states; (d) calculating estimates of phase or amplitude of the largeband signal for each frequency band in a set of multiple frequency bands of the largeband signal, which estimates are based on a portion of the measurements, which portion of the measurements is taken when the RFID tag is in the first reflective state or in the second reflective state and is not taken when the RFID tag is transitioning between the first and second reflective states; (e) transmitting, from the RFID tag, a preamble to an RFID communication; (f) computing, for each specific time segment in a set of time segments, a correlation between (i) a subset of the measurements, which subset was taken during the specific time segment, and (ii) data that encodes the preamble and that was stored in memory; (g) identifying, as a calculated window for the preamble, a time segment for which the correlation is at least as great as that for any other time segment in the set of time segments; (h) determining, based on a switching rate of the RFID tag and on timing of the calculated window for the preamble, a set of transition time intervals in which the RFID tag is transitioning between the first and second reflective states; and (i) disregarding the set of transition time intervals, when estimating a one-dimensional, two-dimensional or three-dimensional position of the RFID tag.
This invention relates to a method for determining the position of a radio frequency identification (RFID) tag using a wideband signal. The method addresses challenges in accurately tracking RFID tags by mitigating interference from state transitions between reflective states, which can distort measurements. The system involves transmitting a wideband signal from one or more transmitters to an RFID tag, which reflects the signal to one or more RF receivers. The receivers measure the reflected signal to detect when the RFID tag is in a first or second reflective state, where the reflection response (phase or amplitude) differs between states. The method calculates phase or amplitude estimates for multiple frequency bands of the wideband signal, using measurements taken only when the tag is in a stable reflective state, excluding transition periods. The RFID tag transmits a preamble, and the system computes correlations between measured data and stored preamble data to identify the preamble's time window. Based on the tag's switching rate and preamble timing, transition intervals are determined and excluded from position estimation. This approach improves accuracy by disregarding data corrupted by state transitions, enabling precise one-dimensional, two-dimensional, or three-dimensional positioning of the RFID tag.
4. A method comprising: (a) receiving, with one or more radio frequency (RF) receivers, a largeband signal; and (b) performing an estimation, which estimation includes (i) calculating estimates of phase or amplitude of the largeband signal for each frequency band in a set of multiple frequency bands of the largeband signal, which estimates are based on measurements taken at a set of receiver antennas of the one or more RF receivers in such a way that measurements taken at each receiver antenna in the set of receiver antennas are separate from measurements taken at each other receiver antenna in the set of receiver antennas, (ii) identifying, based on bandwidth of the largeband signal, a maximum distance and a minimum distance, (iii) identifying, based on the estimates of phase or amplitude, a set of most likely candidate distances, in such a way that (A) each candidate distance in the set of candidate distances is a discrete value, is greater than or equal to the minimum value and is less than or equal to the maximum value, and (B) the set of candidate distances includes, for each specific receiver antenna in the set of receiver antennas, multiple candidate distances between the specific receiver antenna and an RF source of the largeband signal, and (iv) estimating, based on the set of candidate distances, a one-dimensional, two-dimensional or three-dimensional position of the RF source.
This invention relates to radio frequency (RF) signal processing for determining the position of an RF source. The method involves receiving a largeband signal using one or more RF receivers equipped with multiple antennas. The received signal is divided into multiple frequency bands, and phase or amplitude estimates are calculated for each band based on measurements taken at each antenna. These measurements are processed independently for each antenna to ensure separation of data. The method then determines a maximum and minimum possible distance to the RF source based on the signal bandwidth. Using the phase or amplitude estimates, a set of discrete candidate distances is identified, where each candidate distance falls within the defined range and represents a potential distance between each antenna and the RF source. The method then estimates the RF source's position in one, two, or three dimensions by analyzing these candidate distances. This approach leverages multi-band signal analysis and independent antenna measurements to improve positioning accuracy. The technique is applicable in wireless communication, radar, and localization systems where precise RF source tracking is required.
5. The method of claim 4 , wherein the estimation includes performing Bayesian inference.
This invention relates to a method for estimating parameters in a technical system, particularly where uncertainty or incomplete data is present. The method addresses the challenge of accurately determining system parameters when direct measurement is impractical or when data is noisy, incomplete, or subject to variability. The core approach involves using probabilistic techniques to refine parameter estimates based on available data and prior knowledge. The method first collects observational data from the system, which may include measurements, sensor readings, or other empirical inputs. This data is then processed to identify relevant features or variables that influence the system's behavior. The method incorporates prior knowledge about the system, such as historical data, theoretical models, or expert insights, to establish initial parameter distributions. A key aspect of the method is the use of Bayesian inference to update these distributions. Bayesian inference allows the method to iteratively refine parameter estimates by combining prior knowledge with new observational data. This probabilistic framework accounts for uncertainty, providing not just point estimates but also confidence intervals or probability distributions for the parameters. The method may involve iterative steps where new data is continuously integrated to improve accuracy over time. The method is particularly useful in fields such as engineering, physics, and data science, where precise parameter estimation is critical but direct measurement is difficult. By leveraging Bayesian inference, the method provides a robust and adaptable solution for handling uncertainty in parameter estimation.
6. The method of claim 4 , wherein the method further comprises modeling that: (a) is based on the estimates of phase or amplitude; and (b) models a probability distribution of the candidate distances as a mixture of gaussians, where each gaussian is centered at a different integer multiple of a wavelength.
This invention relates to signal processing techniques for estimating distances in systems where measurements are affected by phase or amplitude variations, such as in radar, sonar, or wireless communication applications. The problem addressed is the challenge of accurately determining distances when measurements are corrupted by noise or multipath effects, leading to ambiguous or unreliable distance estimates. The method involves modeling the relationship between observed phase or amplitude measurements and candidate distances. The modeling process is based on estimates of phase or amplitude derived from the received signals. A key aspect of the method is the use of a probabilistic approach to represent the uncertainty in distance estimates. Specifically, the method models the probability distribution of candidate distances as a mixture of Gaussian functions. Each Gaussian component in the mixture is centered at a different integer multiple of the signal's wavelength, accounting for the periodic nature of phase or amplitude measurements. This approach helps resolve ambiguities in distance estimation by capturing the likelihood of multiple possible distance values that correspond to different phase cycles or amplitude variations. The method improves the accuracy and reliability of distance measurements in noisy or multipath environments.
7. The method of claim 4 , wherein the estimation includes solving a Hidden Markov Model.
A system and method for analyzing sequential data involves estimating hidden states from observed data to improve decision-making or predictive accuracy. The method addresses challenges in scenarios where direct observation of underlying states is impractical, such as in speech recognition, bioinformatics, or financial forecasting. By modeling the probabilistic relationships between hidden states and observable outputs, the system infers the most likely sequence of hidden states over time. This approach leverages statistical techniques to handle uncertainty and noise in the observed data, enhancing the reliability of predictions or classifications. The method includes solving a Hidden Markov Model (HMM), which involves defining transition probabilities between states, emission probabilities of observations, and initial state probabilities. The HMM is trained using algorithms like the Baum-Welch method to optimize model parameters based on observed data. During inference, techniques such as the Viterbi algorithm or forward-backward algorithms are applied to determine the optimal sequence of hidden states. The system may integrate additional data preprocessing steps, such as feature extraction or normalization, to improve model performance. The method is particularly useful in applications requiring real-time or near-real-time analysis, where accurate state estimation is critical for decision support or automation.
8. The method of claim 4 , wherein at least a subset of the measurements are gaussian related to each other over time.
This invention relates to a method for analyzing time-series data where measurements exhibit Gaussian relationships. The method involves collecting measurements over time and identifying subsets of these measurements that follow Gaussian distributions or exhibit Gaussian-related statistical properties. By analyzing these relationships, the method enables improved data modeling, prediction, or anomaly detection in applications such as sensor networks, financial time series, or biomedical signal processing. The Gaussian relationships may be used to reduce noise, enhance feature extraction, or improve the accuracy of statistical models. The method may also include preprocessing steps to normalize or filter the data before identifying Gaussian-related subsets. The approach leverages the inherent statistical properties of Gaussian distributions to enhance the reliability and interpretability of time-series analysis.
9. The method of claim 4 , wherein the estimation includes solving a maximum a posteriori problem to estimate a Hidden Markov Model.
This invention relates to statistical modeling and estimation techniques, specifically for estimating parameters of a Hidden Markov Model (HMM). Hidden Markov Models are widely used in various fields such as speech recognition, bioinformatics, and financial forecasting, where the goal is to infer hidden states from observable data. A key challenge in applying HMMs is accurately estimating the model parameters, which often involves solving complex optimization problems. The invention addresses this challenge by using a maximum a posteriori (MAP) estimation approach to solve for the HMM parameters. The MAP framework incorporates prior knowledge about the model parameters, improving estimation accuracy, especially in cases with limited or noisy data. The method involves defining a probabilistic model that includes both the likelihood of observed data given hidden states and a prior distribution over the model parameters. By maximizing the posterior probability of the parameters given the observed data, the method refines the estimates to better reflect the underlying system. The technique is particularly useful in applications where traditional estimation methods, such as maximum likelihood estimation, may produce unreliable results due to insufficient data or model complexity. By leveraging prior information, the MAP-based approach enhances robustness and precision in parameter estimation. The invention can be applied in any domain where HMMs are used, providing a more reliable way to infer hidden states from observable data.
10. The method of claim 4 , wherein the estimation involves solving a maximum a posteriori problem by performing approximate inference, which approximate inference includes a particle filter.
This invention relates to a method for estimating parameters or states in a system using probabilistic techniques, specifically addressing challenges in systems where direct measurement is difficult or noisy. The method involves solving a maximum a posteriori (MAP) problem, which is a statistical approach to finding the most probable values of unknown variables given observed data. To handle computational complexity, the method employs approximate inference, which simplifies the problem while maintaining accuracy. The approximate inference process includes a particle filter, a sequential Monte Carlo technique that uses a set of weighted samples (particles) to represent the probability distribution of the unknown variables. The particle filter iteratively updates these particles based on new observations, allowing the system to track changing states over time. This approach is particularly useful in dynamic systems, such as robotics, sensor networks, or financial modeling, where real-time estimation is required. The method improves upon traditional inference techniques by reducing computational overhead while maintaining robustness to noise and uncertainty. The particle filter component ensures that the estimation remains accurate even with limited or imperfect data.
11. The method of claim 4 , wherein the estimation includes: (a) identifying (i) a set of ellipses that includes, for each specific receiver antenna in the set of receiver antennas, ellipses that each have foci, which foci are the specific receiver antenna and a transmitter antenna, and (ii) a set of most likely points of intersection between ellipses in the set of ellipses; or (b) identifying (i) a set of ellipsoids that includes, for each specific receiver antenna in the set of receiver antennas, ellipsoids that each have foci, which foci are the specific receiver antenna and a transmitter antenna, and (ii) a set of most likely points of intersection between ellipsoids in the set of ellipsoids.
This invention relates to wireless positioning systems, specifically methods for estimating the location of a transmitter antenna using a set of receiver antennas. The problem addressed is accurately determining the transmitter's position when signal measurements (e.g., time of arrival or received signal strength) are subject to noise or multipath interference, leading to positioning errors. The method involves estimating the transmitter's location by analyzing geometric intersections. For two-dimensional positioning, the system identifies a set of ellipses, where each ellipse has foci at a specific receiver antenna and the transmitter antenna. The most likely points of intersection between these ellipses are then determined to estimate the transmitter's position. For three-dimensional positioning, the system uses ellipsoids instead of ellipses, again identifying the most likely intersection points to refine the location estimate. This approach leverages geometric properties to improve accuracy in noisy or multipath environments, reducing reliance on direct signal measurements alone. The method can be applied in various wireless applications, including cellular networks, IoT devices, and asset tracking systems.
12. The method of claim 4 , wherein the estimation includes: (a) identifying a set of ellipses that includes, for each specific receiver antenna in the set of receiver antennas, ellipses that each have foci, which foci are the specific receiver antenna and a transmitter antenna; (b) identifying a set of most likely points of intersection between ellipses in the set of ellipses; and (c) estimating, based on the measurements, a first subset of n intersection points in the set of most likely intersection points, where the sum of pairwise distances between intersection points in the first subset is less than or equal to the sum of pairwise distances between intersection points in any other subset of n intersection points in the set of most likely intersection points.
This invention relates to wireless positioning systems, specifically methods for estimating the location of a transmitter using measurements from multiple receiver antennas. The problem addressed is the challenge of accurately determining the transmitter's position when signal measurements (e.g., time of arrival, angle of arrival) are subject to noise and multipath interference, leading to imprecise location estimates. The method involves analyzing a set of ellipses, each defined by a receiver antenna and a transmitter antenna as foci. For each receiver antenna, multiple ellipses are generated based on signal measurements, representing possible transmitter locations. The method then identifies the most likely intersection points between these ellipses, which correspond to potential transmitter positions. From these intersections, a subset of n points is selected such that the sum of pairwise distances between them is minimized. This subset is considered the most probable transmitter location, as it optimizes geometric consistency among the measurements. The approach improves positioning accuracy by leveraging geometric constraints and reducing the impact of measurement errors through an optimization process. This is particularly useful in scenarios where traditional trilateration or triangulation methods yield unreliable results due to noisy or sparse data.
13. The method of claim 4 , wherein: (a) the estimation involves performing an approximate inference that comprises a particle filter; and (b) the particle filter includes updating particles by sequential importance sampling, to update for new measurements of the largeband signal.
This invention relates to signal processing, specifically to methods for estimating parameters of a largeband signal in the presence of noise or interference. The problem addressed is the accurate and efficient estimation of signal parameters when traditional methods may fail due to high noise levels or complex signal structures. The method involves performing approximate inference using a particle filter, a computational technique for implementing a recursive Bayesian filter by Monte Carlo simulations. The particle filter operates by representing the probability distribution of the signal parameters using a set of weighted random samples, called particles. These particles are updated sequentially as new measurements of the largeband signal are received. The updating process employs sequential importance sampling, a method that adjusts the weights of the particles based on the likelihood of the observed measurements. This ensures that the particle set remains representative of the true posterior distribution of the signal parameters, even as the signal evolves over time. The approach is particularly useful in scenarios where the signal dynamics are nonlinear or non-Gaussian, making traditional estimation techniques less effective. By leveraging the particle filter, the method provides a robust and flexible framework for tracking and estimating signal parameters in challenging environments, improving accuracy and reliability in applications such as communications, radar, and sensor networks.
14. The method of claim 4 , wherein the estimation further comprises replacing each outlier estimate, in a set of outlier estimates of position, with an interpolated estimate of position, which interpolated estimate is interpolated from a previous position.
This invention relates to improving position estimation in systems where outliers or erroneous measurements can degrade accuracy. The method addresses the problem of unreliable position estimates caused by outliers, which can occur in various applications such as navigation, tracking, or sensor networks. The solution involves identifying outlier estimates within a set of position estimates and replacing them with interpolated values derived from previous valid position data. This interpolation ensures continuity and reliability in position tracking by smoothing out erroneous spikes or deviations. The interpolation process may use linear, polynomial, or other suitable methods to estimate the corrected position based on prior valid measurements. This approach enhances the robustness of position estimation systems by mitigating the impact of outliers, leading to more accurate and consistent tracking over time. The method is particularly useful in environments where sensor noise, interference, or other factors introduce errors into position measurements. By dynamically replacing outliers with interpolated values, the system maintains high accuracy without requiring additional hardware or complex algorithms.
15. A method comprising: (a) wirelessly transmitting a largeband signal in such a way that the largeband signal travels from one or more transmitters to a backscatter node, reflects from the backscatter node and travels to one or more RF receivers; (b) taking measurements of the largeband signal at a set of receiver antennas of the one or more RF receivers; (c) identifying, based on the measurements, (i) when the backscatter node is in a first reflective state and when the backscatter node is in a second reflective state, the first reflective state differing from the second reflective state in that a reflection response of the backscatter node is different in the first reflective state than in the second reflective state, which reflection response is phase or amplitude of RF reflection from the backscatter node as a function of RF radiation incident on the backscatter node, or (ii) when the backscatter node is transitioning between the first and second reflective states; (d) calculating estimates of phase or amplitude of the largeband signal for each frequency band in a set of multiple frequency bands of the largeband signal, which estimates are based a portion of the measurements, which portion of the measurements is taken when the backscatter node is in the first reflective state or in the second reflective state and is not taken when the backscatter node is transitioning between the first and second reflective states; (e) identifying, based on bandwidth of the largeband signal, a maximum distance and a minimum distance; (f) identifying, based on the estimates of phase or amplitude, a set of most likely candidate distances, in such a way that (i) each candidate distance in the set of candidate distances is a discrete value, is greater than or equal to the minimum value and is less than or equal to the maximum value, and (ii) the set of candidate distances includes, for each specific receiver antenna in the set of receiver antennas, multiple candidate distances between the specific receiver antenna and an RF source of the largeband signal; and (g) estimating, based on the set of candidate distances, a one-dimensional, two-dimensional or three-dimensional position of the RF source.
This invention relates to wireless positioning systems using backscatter nodes to determine the location of an RF source. The system addresses challenges in accurately locating RF sources by leveraging backscatter nodes that modulate their reflective properties between two distinct states, altering the phase or amplitude of reflected signals. The method involves transmitting a wideband signal from one or more transmitters to a backscatter node, which reflects the signal to one or more RF receivers. Measurements of the reflected signal are taken at multiple receiver antennas to identify when the backscatter node is in its first or second reflective state or transitioning between them. Phase or amplitude estimates of the wideband signal are calculated for multiple frequency bands, excluding measurements taken during state transitions. Based on the signal bandwidth, a maximum and minimum distance range is determined. The system then identifies discrete candidate distances between each receiver antenna and the RF source, ensuring they fall within the defined range. These candidate distances are used to estimate the RF source's position in one, two, or three dimensions. The approach improves positioning accuracy by exploiting the backscatter node's reflective state changes and wideband signal analysis.
Unknown
June 30, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.